Hacker News new | past | comments | ask | show | jobs | submit login
Become a Signal Sustainer (signal.org)
69 points by feross on Dec 1, 2021 | hide | past | favorite | 57 comments



What pushed me away from Signal, after advocating for it for years (in fact deleting WhatsApp for more than a year; WhatsApp IS messaging in my country. Period), trying to get friends and family on it, is that Signal really doesn’t care. It doesn’t care how people will easily use it, like it. When there were issues like endless notification hell for cross platform connections they were busy doing crypto. They talked about ID and it’s nowhere to be seen. Messages delays, non-delivery were widespread issues and there were not even a comment by Signal staff, not a KB article.

Also the amount of hostility and derision they show to their own users and possible contributors at places like Github is what finally did it for me.

At this point I have no patience left for Signal and after years my friends and contacts were pleasantly surprised to find me back on WhatsApp as the primary messaging app. Many of them had signal just to communicate with me that they uninstalled.

Most of them really tried using it (trying to convince their own circle) and found it essentially unusable compared to WhatsApp and Telegram.


So different from my experience. Reliability same/better with signal compared to most any other messaging systems save apple messages.

Less technical friends and family say it looks like text messenger/apple messenger.

Haven't seen the hostility. Any examples?


Maybe I'd care if they didn't spend a year of donation-funded development time on Mobilecoin integration.


I don't think we should puritans in this case. Signal is an important force against pushes for building the panopticon.


I agree, its really frustrating since that time and effort could've gone into messaging/communication features


Funny, I was just thinking I'd much rather they launched something like Mobilecoin instead of depending on donations.


I thought similar, but a proper DAO. Security might be an issue but it would be ideal for this project.


> Maybe I'd care if they didn't spend a year of donation-funded development time on Mobilecoin integration.

Well, the donation system itself does not allow mobilecoin as a payment method even when the option to enable it is available. Perhaps they are dropping this cursed thing for good.


Isn't this an alternative to that funding strategy?


I donated to Signal Foundation a while back and tried to get my employer donation match but no one from Signal ever responded to their information requests. Oh well... their loss


The Signal Technology Foundation is in Benevity. My employer matches 100%.


Yeah but I think what OP means is that the beneficiary non-profit needs to confirm the donation to receive the company match. Benevity and other services that companies outsource their giving and matches to are required to do due diligence to confirm the original donation before they can send the company match.


That explains the badge thing! Time to donate. Signal is not perfect and has a few things that make me really mad, but it is the best grandma-compatible IM we have.

I also hope they start listening to their users or integrating some PRs already contributed on GitHub.


Grandma-compatible secure IM.

I don't want my private conversations leaking some time in the future and I trust Signal way more than I trust Whatsapp in this regard.


LOVE how they don't autorenew subscriptions. Despite the MOB fiasco, I think the people behind this still have good intentions.


will the books be open sourced? e.g. how much was donated vs how much it costs to run signal? Whom gets any excess? If it's truly non-profit, why not form a 503c that I can get a tax deduction for? That last part is what is required for larger donors to come on board.


They are a non profit, they just havent released their numbers in a few years https://projects.propublica.org/nonprofits/organizations/824...


Leave it to Signal to keep things hidden for long periods of time (like the server code while it was building some cryptocurrency). Seriously, Signal is not an open or transparent organization by any measure. Ironically, it’s an authoritarian and user hostile platform when examined at a deeper level. Instead of allowing other clients or allowing chat backups (still not possible on iOS) or not exposing your phone number by default to everyone else, it hammers on both trivial and very hard matters that aren’t what a typical user wants right now.


Check signal's homepage. At the bottom, it says "Signal, a 501c3 nonprofit". Is that what you mean?


This is a wonderful opportunity for all of us to put our money where our mouths are.


The last thing I want is another centralized and monolithic platform that requires a phone number, exposes phone numbers to everyone, and takes weird decisions on non-disclosure (be it source code of parts of the platform or the financial information as a non-profit) to be the one everyone promotes and uses. Based on the priorities Signal follows, which do not match mine,I don’t see value in this. The protocol may be good, but the platform is not.


I don’t use it very often (not enough people from my circle there) but I went straignt to that screen and donated. We gotta keep this option live and open for us.


Signal did made some significant progress across IM space due to Axolotl/Signal Protocol which has to be noted, but I'm not going to use or support a yet another closed silo when there are open and federated networks around that work well.


Yes or if they'd open up the network to third party clients. Then I'd support them.

I prefer matrix but it doesn't feel as ready for prime time. I doubt the masses will ever adopt it. Whereas signal is really just like WhatsApp.

But I want to use whatever client I want. Sometimes I'm in a shell session on FreeBSD, sometimes on a computer. One of many. Rarely on my phone. Signal is too mobile-first for me. Third party clients could fill that gap.

What I'd love to see is for them to operate it as an open network, not a closed network with an open source app. I wouldn't mind at all if it cost money. I rather pay cash than with my data.


Moxie makes some good arguments for having the server centralized, which is evidenced in Matrix moving/developing insanely slow due to it's decentralized nature.


Matrix is not really that slow IMO. They've done a lot over the last years. Their E2E encryption is a ton better. And what's there works well. And what has signal delivered in the last 2-3 years? Not much either. The product is just finished.

I'm sure it makes development a lot easier but it also makes signal a lot less viable. I'm not going to convince my friends to move into yet another walled garden we have to move out of again at some stage.


Also I wanted to add: matrix is also trying to do a lot more than signal is. Signal is not substantially different than for example WhatsApp, only with more responsible owners.

Matrix is aiming at being a completely open network standard, identity service (can use a phone number but doesn't have to!), multiuser conferencing system etc, in the decentralised spirit of the early internet. In that sense it's more comparable to MS Teams (which is moving fast in development but MS is rolling in it and they're making it slower and buggier with every version recently)

It's not perfect but considering their scope is much bigger, they're moving fast enough IMO.

Moxie has mentioned before that clients not supporting certain features was a burden but this is not something that Matrix seems to be having much delay from. If a client doesn't support pictures, audio messages or whatever they just don't show them.

Ps when I said matrix E2E encryption is a ton better I mean it's much better than it used to be 2 years ago. Not better than signal, which is technically better but more limited in terms of number of clients.


Fair points and I agree. And I love Matrix as a concept, it just has been really rough around the edges in practice (been on and off since 2017 or so and regular user since 2019). As I see it Signal is superior to Matrix in the (Whatsapp|Signal|FBMessenger|Telegram|whatever) space, but Matrix has great potential across the board. It just irks me when people suggest using Matrix (Element) as a replacement to Signal - the use cases are very different and the UI/UX and overall user friendliness are very different - I would never try to get my aunt on Matrix, whereas on Signal the onboarding and use is just the same as any other regular messaging app, except with top notch encryption. And I've been using Signal/TextSecure since 2013-2014 (granted I didn't have many (any) contacts using it back then).

And overall I'm happy to see continuous progress with Matrix, it's just slow IMO. But obviously you need to consider what's actually involved under the hood - where Signal has a much easier job due to its centralization.

Personally I wish both projects great success!



Yeah I know. I currently use the matrix bridge and it works fine. But technically it's not allowed and they could ban it. Right now I only have 2 contacts that use Signal exclusively but they're family members so I really need it.


Right, that's problematic as well - Signal doesn't even have an official client for my smartphone.


They really, really have to get rid of phone numbers. It's miserable and unsafe.


What's less miserable and safer?

What prevents not just message privacy but good metadata resistance? What prevents mapping out your social network? What prevents an attacker from create 100M accounts and flooding all current users with invites and/or messages?

I get that tying signal to a sim is less than perfect, but what is better?


Don't they have a cryptocurrency to fund them instead, now? I mean, I get it, asking for money means more funding than not asking for money, but it combines in poor taste, in my opinion.

I suppose at least they haven't gone in on NFTs, so there's still hope.


how they managed to pay their employees in the bay area up until then?

donations? i don't buy it

where can i find how they manage their finance?


here is the last time they published their numbers

https://projects.propublica.org/nonprofits/organizations/824...


2018 - 609k of revenue (100% from investment)?

2019 - 19M of revenue

--

2018 - 450k to AWS (indicates they have an infrastructure and users)

2019 - 887k to AWS

so their invoice for AWS only just doubled and yet their revenue increased by 3019.87%

very fishy activity

i'd love to be able to see their cryto wallets, i suspect some money laundering activity


> yet their revenue increased by 3019.87%

They received a $50 million donation from the co-founder of WhatsApp. Elon Musk also claims to have made a donation, but that is unconfirmed.


I've pretty much stopped using Signal since they said they're not going to port it to M1. Probably would have done this otherwise.


I really don't get their hostility to low hanging fruit. They continue to only publish a Debian package, if you're on any other distribution you need to use a community built Flatpak. They could improve the security of everyone by just building the same Flatpak themselves!


The Signal xbps package works fine, and Flatpaks are security hell.


Signal runs without issue on my M1 Macbook Air.


I assume they mean native ARM support rather than depending on Rosetta 2, which from what I can tell [1] is still the case.

[1] https://doesitarm.com/app/signal/


I also think it was more about ARM.....I dont like using it since it wont work on my ARM Chromebook, there are also the huge battery drain issues on Android and mediocre video/audio calling


Well they'll have to port it eventually. Rosetta isn't going to be around forever.


It runs under Rosetta 2. Not natively.


While that certainly squares with my mental model, do you happen to have a link to that decision?


Sorry, yes. There's multiple sources, but a lot of it collected here: https://github.com/signalapp/Signal-Desktop/issues/4461

They keep shutting down tickets about it, refuse to review a PR that implemented it for them, and won't even tick the box on the Signal iPad app that would make it available on M1 Macs (where side-loaders have already confirmed for them that it works perfectly.

90% of my Signal usage was on desktop, so this has left such a bad taste that I'd rather use almost anything else now.


> won't even tick the box on the Signal iPad app that would make it available on M1 Macs

Isn’t that box ticked by default? I think what Signal has done is gone in and deliberately unticked it to make you use the Intel Mac app in Rosetta rather than the iPad app.

Edit: yup

> After you sign the updated developer agreement, the App Store automatically makes compatible iOS apps available to users of a Mac with Apple silicon. However, if you’re already planning to ship a macOS version of your app, or if your app doesn’t make sense on the Mac, you can change your app’s availability in App Store Connect.


Wow. That's even more user-hostile then.


I can see it making sense if you're a cross platform shop and a high quality Mac app is something that you're prioritizing. You might not want people to inadvertently happen on an iPad app that will run on the Mac but is worse than the Mac app.

But it doesn't sound like Mac is a big priority here, so who knows why they would do that.


> tick the box on the Signal iPad app that would make it available on M1 Macs

This annoys me the most. I don't see the downside of them enabling it. It would bring the Mac client into feature parity with the IOS client and eventually reduce complexity. It seems like it would be an easy win-win.


The reason for this is that the iOS and macOS security models for attachments are not the same, and so additional work has to be done before it could work. It’s not just click and go.

(At least this is what I read months ago)


Is there a substantial benefit to running something like Signal on M1 natively vs. through Rosetta? I don't have an M1 Mac so I don't a feel for this issue but my understanding was that between Intel and M1+Rosetta, x86-64 native applications run approximately the same.


Battery life, especially with regards to Electron.


Not the person you replied to, but a GitHub issue thread [1] seems quite close.

I find it somewhat confusing and potentially troubling that (from a few minutes of searching so correct me if I’m wrong) a full-featured ARM-native fork seems to have been created in the community, but the team apparently doesn’t want to go the extra mile of designing the build/CI system to incorporate this as an official release.

[1] https://github.com/signalapp/Signal-Desktop/issues/5612


That's why I said it squares with my mental model. IIRC they have been super hostile to any 3rd party clients, which is (likely) what "created in the community" means to them

I also have a ton of bitterness about their build system since it seems every few months they make their gradle and codesigning process more and more hostile to self-building. And I appreciate that is kind of an edge case but if they ever want people to chip in and try to fix the seemingly unbounded MMS (and now SMS!) bugs, they'd be well served to have less wizardry




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: